Insurance-Canada.ca - Where Insurance & Technology Meet

AI: The Computing We Always Wanted?

With all of the discussions about InsurTech, Disruption, Innovation, Artificial Intelligence (AI) may  be poised to slide in seamlessly and radically change how we do business.  Are we realizing what we thought computing should be?

Where is AI now in insurance?

We are already seeing implementations of AI.  Lemonade, which is bringing a new model to residential insurance, is utilizing AI to help adjudicate claims quickly for a large portion of the insureds.

And there are a number of opportunities in the pipeline…

Think about the complexity of risks that we are facing, AI could provide more comprehensive information for better, potentially dynamic, risk management and underwriting.

In reviewing some of the applications on display at the Consumer Electronics Show (CES), Anthony O’Donnell, Executive Editor, Insurance Innovation Reporter, made an interesting, observation.  “AI (Artificial Intelligence) is unmistakably on the march in the insurance industry and will become an important competitive weapon over a shorter timeframe than many might have predicted” (emphasis supplied).

We have all experienced ‘advisors’ and ‘chatbots’…

Utilities and government entities have introduced machines to guide inquiries and simple transactions.  As well,insurance companies and brokers have implemented intelligent phone systems to handle routine requests.

At the CES, O’Donnell interviewed Donald Light, Celent’s Director North America Property Casualty Practice.  Light sees IPAs (Intelligent Personal Assistants) as providing a step above other  personal assistants such as chatbots and robo-advisors.

The difference is that new devices, such as AWS’ Alexa, not only supply data, but collect data to provide information on what people are doing and thinking.

Deep Learning goes deep

One of the reasons why AI is so important to insurance is that AI is just as important to most other sectors.  The base products underlying voice recognition (the smart interface for AI) are applicable to a full range of scenarios.  Think of all the interesting questions that Apple’s Siri gets.

Last fall, Fortune documented developments in AI and the implications.  Fortune found that users “are increasingly interacting with our computers by just talking to them”.  Baidu, the Chinese counterpart of Google, reported that  “customers have tripled their use of its speech interfaces in the past 18 months.”

In addition, there are advances in image recognition as well as voice.  At present, Baidu, Google, Microsoft, and Facebook are applying image recognition to managing photos.  However, image recognition can extend to other sectors, including medicine (reading XRays, MRIs, etc.), and enabling self driving vehicles, robotics and more.

Fortune notes that these developments are part of the AI family, using techniques that are known as ‘deep neural networks’.  The most significant element is  “that no human being has programmed a computer to perform any of the stunts described.”

Put simply, Computers can now teach themselves.

Why is AI moving so quickly?

I have a theory… AI is what we all thought IT would be.  As much as we were intrigued by computers large and small, as much as the Internet supplied the capabilities to access data world wide, as much as we found old and new friends online, these functions were incremental.

The promise of computing has always been to simplify routine data and transactional functions.   Beyond that, we can  expect that machines will learn to help us improve decision making and create a platform to support innovation.  Put simply, AI will help us think.

Put simply, when we boomers saw the Jetsons, that was what we wanted.  And AI is showing that we can have it.

We will be focussing on AI …

We plan to look into AI and insurance in this space on a regular basis.

Also, we have several sessions focussing on AI and its pillars of support at #ICTC2017, including: